|
| ▲ | JacobAsmuth a day ago | parent | next [-] |
| I think this is a completely different class from Luna and is much more akin to 6.1 Sol in terms of capabilities and general intelligence. |
| |
| ▲ | redox99 a day ago | parent [-] | | Maybe, although google tends to benchmaxx. If it's actually sol for free that's kinda crazy good. |
|
|
| ▲ | WarmWash a day ago | parent | prev | next [-] |
| 3.8 flash was easily the best general everyday person model yet though |
| |
| ▲ | notatoad a day ago | parent | next [-] | | 3.8 flash was never served to free users on google though? | | | |
| ▲ | Gareth321 a day ago | parent | prev [-] | | Maybe I’m spoiled but I found it repulsively dumb. Some of the highest rates of hallucinations since GPT-3. Incredibly lazy. Heavily biased to parametric knowledge and speed over accuracy. I would rather have no AI at all than use that. |
|
|
| ▲ | seizethecheese a day ago | parent | prev [-] |
| It's definitely not the case that small models only do well in benchmarks. We're working on a battle system on top of OpenCode, so have data for real people doign real work. Right now DeepSeek Flash 4.1 Flash has a 64% win rate against GPT 6.1 Sol. Small data still, but it has 33 wins vs. 18 losses, which is confidence interval of 52% - 78% win rate. https://agentduel.com/leaderboard/models/deepseek/deepseek-v... |
| |
| ▲ | redox99 a day ago | parent [-] | | That's coding though. I was talking about general usage. | | |
| ▲ | seizethecheese a day ago | parent [-] | | Okay, makes sense, though I don’t see a huge difference in quality between Luna and Astra for basic chat usage. |
|
|
